Herpes simplex virus (HSV) is a highly prevalent, clinically important pathogen that establishes lifelong latent infections in sensory neurons, leading to recurrent infections. During latency, productive HSV infection is undetectable, but there is transcription of Latency Associated Transcripts (LATs), including viral microRNAs (miRNAs) from one region of the viral genome. The role of the LATs remains unclear, but some studies have linked this transcription to neuronal survival. In addressing this problem, our previous studies found that if Dicer, an essential enzyme in miRNA biogenesis, is knocked out in most infected neurons, 60% of HSV genomes are lost during latency, but this was not due to the loss of neurons. In other words, persistence of latent virus, but not neuronal survival, was compromised in the absence of Dicer. We now follow these findings in two ways: First, we knocked out Drosha, another enzyme required for most miRNA biogenesis; second, we used a modified genetic model in which the knock out of Dicer is driven by the HSV LAT promoter, which should reach a greater proportion of latently infected neurons, with the caveat that this virus does not transcribe the major LATs. We found that knock out of Drosha impacts HSV latency, but less than Dicer, with 30% of viral genomes being lost. Further supporting this complexity, when knock out of Dicer was driven by the LAT promoter, loss of viral genomes was around 60%, as in our original model, but this was now accompanied by the loss of many neurons. Together these results suggest that Dicer has roles in HSV latency that go beyond canonical miRNA production and that loss of Dicer reveals an essential role for LATs in supporting the survival of latently-infected neurons.